Closed strengejacke closed 1 year ago
I think this is due to https://github.com/easystats/insight/pull/514, where find_formula()
for rma was implemented, and methods for other meta-analysis packages redirected to find_formula.rma()
. I reverted this change for meta-plus package.
The new issue I see now is:
library(metaplus)
#>
#> Note: The second parameter is the study standard error not the standard error squared as in the metafor package.
library(dplyr)
#>
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#>
#> filter, lag
#> The following objects are masked from 'package:base':
#>
#> intersect, setdiff, setequal, union
data(mag, package = "metaplus")
dat <- mag %>% rename(estimate = yi, std.error = sei)
set.seed(123)
df <- statsExpressions::meta_analysis(
data = dat,
type = "robust",
random = "normal"
)
#> Error in seq_len(nrow(mf)): argument must be coercible to non-negative integer
Created on 2023-03-15 with reprex v2.0.2
ok, also related to that PR. Works now, no revdep-issue here.
library(metaplus)
#>
#> Note: The second parameter is the study standard error not the standard error squared as in the metafor package.
library(dplyr)
#>
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#>
#> filter, lag
#> The following objects are masked from 'package:base':
#>
#> intersect, setdiff, setequal, union
data(mag, package = "metaplus")
dat <- mag %>% rename(estimate = yi, std.error = sei)
set.seed(123)
statsExpressions::meta_analysis(
data = dat,
type = "robust",
random = "normal"
)
#> # A tibble: 1 × 14
#> term effectsize estimate std.error conf.low conf.high
#> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 Overall meta-analytic summary estimate -0.746 0.234 -1.26 -0.343
#> statistic p.value weight conf.level method
#> <dbl> <dbl> <dbl> <dbl> <chr>
#> 1 -3.20 0.000501 NA 0.95 Robust meta-analysis using 'metaplus'
#> conf.method n.obs expression
#> <chr> <int> <list>
#> 1 Wald 16 <language>
Created on 2023-03-15 with reprex v2.0.2
Thanks, @strengejacke! Glad it turned out to be a non-issue.
Not sure which change in insight caused this issue, maybe we can even avoid breaking code?